Project Summary: Experimental model organisms such as mice, rats, and Drosophila provide valuable insights into human diseases. This is because we can manipulate genetic and environmental factors in experimental populations. Biological constraints limit our ability to manipulate genetic systems, and infer causation (relative to industrial experimental settings). Our proposal addresses unique challenges due to these constraints in experimental genome populations used to study human diseases. We will develop statistical design and analysis methods that will reduce experimental cost, make efficient use of existing resources, and better infer causation when we have incomplete control over the assignment of genetic factors to individual organisms. We focus on three types of collections of genetic perturbations: (1) experimental crosses (derived from crossing two or more strains such as backcross, intercross, and recombinant inbred line populations), (2) introgression line populations (such as consomic strains, and genome-tagged mice), and (3) non- randomized genome populations (such as inbred strain collections). Our specific aims are: (1) To develop efficient and robust genotyping strategies in experimental crosses (2) To develop statistical design and analysis tools for introgression line populations. (3) To develop statistical methods for detecting causal elements in genetic association studies in experimental organisms. Relevance: Scientists discover genetic elements contributing to human disease susceptibility by studying experimentally-constructed animal populations (such as mice). We will develop statistical methods to construct and use these populations efficiently. Our methods will reduce experimental cost, and improve detection of causal genetic elements contributing to disease susceptibility.